Glioblastomas (GBM) use different and distinct anatomical structures to invade the normal brain parenchyma. Although a lot of attention has been given to the infiltrating invasiveness of GBM percolating through the extracellular matrix, still very little is known about perivascular invasion in which GBM cells travel on the outside of blood vessels in the perivascular space dispersing significant distances into the brain parenchyma. This phenotype is specific for high-grade gliomas. Perivascular invasion requires at least two cell types: the endothelial cells that form the vascular tubes and the tumor cells. This project will test the hypothesis that perivascular GBM dissemination is orchestrated by the interaction of astrocytoma-specific factors and brain-specific blood vessel components, and that the onset of angiogenesis restricts this invasive process. One such factor that impacts the balance of angiogenesis and perivascular invasion is MMP-9, which will be the focus of Aim 2. AIM1: REVEAL THE CRITERIA FOR PERIVASCULAR GBM DISSEMINATION. AIM 1.1: Determine to which extent astrocyte-specific criteria are necessary AIM 1.2: Determine to which extent brain-specific vascular aspects are essential, and whether angiogenesis impacts perivascular invasion. AIM2: STUDY METALLOPROTEINASE MMP-9 IN GBM DISSEMINATION We hypothesize that MMP-9 has two functional roles in GBM progression dependent on its location. Tumor cells at the invading front use MMP-9 to cleave the extracellular matrix and infiltrate into the brain matrix as single cells. MMP-9 in host cells (macrophages, endothelial cells, and bone marrow), however, initiates angiogenesis and thereby restricts perivascular GBM invasion. AIM2.1: Identify the exact cell sources of MMP-9 production in GBM. AIM2.2: Reveal function of host cell-produced MMP-9 in angiogenesis and invasion. AIM2.3: Elicit function of tumor cell-produced MMP-9 in angiogenesis and invasion.